I thought a lot about the future of various fields, including rad onc, prior to choosing rad onc in med school. Now, at the mid-point of residency, I can see that a lot of my big-picture worries about the field were wrong.
The big-picture view of the future of the field depends on two related factors: (1) job market forces such as supply/demand which are likely to ebb and flow over the next 40 years, and (2) fundamental changes in how cancer is treated.
First, addressing the job market forces, it is true that right now there has been a recent increase in the number of residents, and if you read the rad onc board, there is some (anecdotal) evidence that this increase has led to a supply-demand imbalance in the job market. There are also some recent changes in RT demand, such as hypofractionation for prostate and breast (hypofractionation means less fractions for a course of radiation to provide the same clinical benefit--radiation oncologists are generally paid per fraction so this means less money in your pocket per patient). These are important considerations, especially for current residents or recent grads, but in my opinion these are relatively minor job market fluctuations in the grand scheme of a 40+ year career in the field. Graduating from a below-average program and wanting to be in a desirable location could be tough due to these factors, but the majority of radiation oncologists will have plenty of work to do and will enjoy a good lifestyle/income. And while there are certain situations where RT is being used less or for shorter fractions, we are in the midst of several emerging areas of RT, including stereotactic treatment in general, which is reimbursed very highly. And in particular, certain areas that might really increase the amount of volume of RT are lung SBRT for early stage lung cancers (lung cancer screening is just starting to pick up), SRS for brain mets, and SBRT for oligometastatic/oligoprogressive disease. In general, I think as the supply/demand balance shifts, the very smart physicians in our field will figure out new things to do and treat, as has happened in many fields within medicine for decades.
Second, (and more to the point of the OP's question), fundamental changes in how cancer is managed are trending toward increased use of radiation therapy. At one point I was concerned that machine learning could make contouring/RT planning obsolete, and at another point I was concerned that immunotherapy or whatever next big systemic therapy there is could make RT obsolete altogether. Both of these are likely untrue. First, contouring is a small part of our workflow, and improvements in contouring efficiency would not change the overall patient volume, in the same way that being able to dictate your notes in primary care would not change the number of PCP's we need--the ultimate demand is driven by the number of patients. Second, better systemic therapy increases the need for RT. We are treating more and more metastatic disease because patients are doing very well on long-term immunotherapy or targeted therapy. Often, this "palliative" treatment is done stereotactically, so requires a lot of technical effort/expertise and is reimbursed at a high rate.
Finally, to address the point above about the salaries in MGMA not being representative, I am just a resident so cannot say for sure, but my anecdotal experience from talking to the senior residents in our (well-respected) program and neighboring programs is that starting salaries are typically ~$300,000 (possibly slightly lower in PP), and all of the residents I know who went into PP were on partnership tracks with generally a doubling of salary at the partner stage. It is true that your salary can increase a lot if you become partner, but it is not true that the older docs are keeping the younger ones out from doing this (again, in my limited anecdotal experience).
